Risk Factors in Dental Erosion

V.K. JARVINEN, I.I. RYTOMAA, and O.P. HEINONEN1

Department of Cariology, University ofHelsinki, Mannerheimintie 172, SF-00300 Helsinki, Finland; and 'Department ofPublic Health, Universityof Helsinki, Haartmaninkatu 3, SF-00290 Helsinki, Finland

Dental erosion and factors affecting the risk of its occurrencewere investigated with a case-control approach. One hundredand six cases with erosion and 100 randomly selected controlsfrom the same source population were involved in the study.All cases and controls were evaluated by the recording of struc-tured medical and dietary histories and by examination of theteeth and saliva. Erosion was classified according to pre-de-termined criteria. The relative importance of associations be-tween factors and erosion was analyzed by a logisticmultivariable model. Adjusted odds ratios (AOR) were esti-mated. There was considerable risk of erosion when citrusfruits were eaten more than twice a day (AOR 37), soft drinkswere drunk daily (AOR 4), apple vinegar was ingested weekly(AOR 10), or sport drinks were drunk weekly (AOR 4). Therisk of erosion was also high in individuals who vomited (AOR31) or exhibited gastric symptoms (AOR 10), and in thosewith a low unstimulated salivary flow rate (AOR 5).

J Dent Res 70(6):942-947, June, 1991

Intrinsic causes of erosion include recurrent vomiting as aresult of psychological disorders, e.g., in anorexia and bulimia(Hellstrom, 1977; Knewitz and Drisko, 1988), or regurgitationof gastric contents because of some abnormality in the gastro-intestinal tract (Ismail-Beigi et al., 1970; Eccles, 1978; Pope,1982; Myllirniemi and Saario, 1985).

One important additional factor in dental erosion is lowsalivary flow, which, naturally, results in inadequate rinsingand buffering of demineralizing acids on tooth surfaces (Dawes,1970; W6ltgens et al., 1985). The roles of salivary calciumand phosphorus in the development of erosion are not yet known(Mannerberg, 1963; Frostell, 1973; Woltgens et al., 1985).

Although many causes of dental erosion are known, theirrelative degrees of importance are not clear, whether from thepoint of view of levels of risk associated with individual ero-sion-causing factors or from the point of view of their effectsin populations. The aim of this study was to investigate factorsrelating to dental erosion, with emphasis on estimation of thequantitative contributions of non-industrial risk factors.

Introduction.

Dental erosion is defined as loss of dental hard tissue by achemical process that does not involve bacteria (Pindborg, 1970).Over the last two decades, interest in erosion has increased(Eccles and Jenkins, 1974; Hellstrdm, 1977; W6ltgens et al.,1985; Linkosalo and Markkanen, 1985; Asher and Read, 1987;Rytomaa et al., 1988; Jirvinen et al., 1988; Sorvari, 1989).In developed countries, the incidence of the previous majordental disease, caries, has declined. More and more patientsand teeth are therefore at risk of developing other dental le-sions. The risk of erosion can increase as a result of changesin dietary habits and when gastric symptoms are exhibited,especially if certain psychological eating disorders are alsopresent (Hellstr6m, 1977; Hurst et al., 1977; Eccles, 1978;Clark, 1985; Negus and Todd, 1986; Knewitz and Drisko,1988).

Dental erosion can have extrinsic or intrinsic causes. Ex-trinsic factors include demineralizing acidic foods-such ascitrus fruits and acidic beverages (Eccles and Jenkins, 1974;Smith and Knight, 1984; Asher and Read, 1987)-and somemedicines-such as effervescent vitamin C preparations,chewable vitamin C tablets (Eriksson and Angmar-Mansson,1986; Meurman and Murtomaa, 1986), and iron tonics (Jamesand Parfitt, 1953). Another extrinsic cause of erosion is acidsin the air breathed. Before effective occupational health mea-sures were adopted, exposure to acids was common in thechemical and metal industries (Miller, 1907; Lynch and Bell,1947; Malcolm and Paul, 1961; ten Bruggen Cate, 1968; Sko-gedal et al., 1977). Nowadays, exposure to extrinsic acid canalso be associated with leisure activities such as frequent swim-ming in chlorinated pool water (Centerwall et al., 1986).

Materials and methods.A case-control study was conducted. One hundred and six

cases of erosion were noted between 1986 and 1987 by dentistsin the Metropolitan Helsinki area. One hundred controls wererandomly selected from the same source population as the cases;a control was the first patient visiting the same dentist on thenext morning following the visit of a case. Both the cases andthe controls were referred to the University Dental Clinic, wherethey were examined. Six controls did not attend for clinicalexamination. Five of the remaining 100 controls sent for ex-amination by their dentists turned out to be erosion cases.Because the erosion cases were detected (according to strictlydefined criteria) among 100 controls in the random samplefrom the source population (the patients of Helsinki dentists),the prevalence of erosion with 95% confidence limits was es-timated from this sample. Interviews with the six controls whorefused examination, and their dentists, did not suggest thatthey were suffering from obvious dental erosion. The five pa-tients removed from the control series were replaced by othercontrols selected at random.

In the case-control series, men and women were about equallyrepresented among both cases and controls. The mean age ofthe cases was 33.1 years (range, 13-73 years). The mean ageof the controls was 36.3 years (range, 17-83 years).

Erosion was classified according to the criteria of Ecclesand Jenkins (1974). The following findings characterized anerosion case: (i) absence of developmental ridges on the enamel,resulting in a smooth glazed enamel surface; (ii) concavitiesin the cervical region on the labial enamel surfaces whosebreadth greatly exceeded their depth, thus distinguishing themfrom cervical abrasion lesions; (iii) edges of amalgam resto-rations raised above the level of the adjacent tooth surface;and (iv) depression of the cusps of posterior teeth, producing'cupping'. In more severe cases, dentin was also involved. Amore detailed description of the lesions seen by us is reportedelsewhere.

942

Received for publication November 23, 1990Accepted for publication February 12, 1991This study was supported by The Finnish Dental Society and Fin-

nish Women Dentists.

RISK FACTORS IN DENTAL EROSION

Structured medical and dietary histories were taken. Em-phasis was placed on past and present gastric complaints suchas vomiting, belching, heartburn, an acid taste in the mouth,stomach-ache, and gastric pain on awakening. Dietary habitswere carefully evaluated by questions relating to special dietsand to the quantity and frequency of consumption of citrusfruits, berries and other fruits, soft drinks, sport drinks, vita-min C, acid sweets, apple vinegar, pickles, and other acidicfoods.

Unstimulated and stimulated (by the chewing of paraffin)saliva samples were collected over five-minute periods. Theunstimulated salivary flow rate was classified as low when itwas 0.1 mL/min or less. For stimulated salivary flow, thecorresponding limit was set at less than 1 mL/min. The acidity(pH) of the saliva was measured with pH electrodes. Bufferingcapacity was measured by the Dentobuff method (Orion Diag-nostica, Finland). Salivary calcium was measured with an atomicabsorption spectrophotometer (Perkin-Elmer, Norwalk, CT),by the method of Trudeau and Freier (1967), in samples thathad been stored in a deep-freeze. Phosphorus was measuredas described by Fiske and Subbarow (1925), and protein bythe method of Lowry et al. (1951).The association between exposure and dental erosion was

first described in terms of unadjusted odds ratios (OR), whichindicate the risk of erosion in exposed (as opposed to non-exposed) subjects, taking no account of the effects of otherfactors. Cases and controls were classified as exposed or non-exposed with respect to extrinsic and intrinsic factors, as in-dicated below and in the "Results" section.The relative importance of associations between erosion and

different risk factors was determined with a logistic multivar-iable model fitted to the data by means of the Glim 3 computerprogram (Nelder and Wedderburn, 1972). For this analysis,the factors were dichotomized (limits are given in Table 4).Confounding factors such as age and gender were included inthe model. Initially, all factors were analyzed. Later, oncemore realistic models had been constructed, non-significantfactors were dropped from the model. Interactions betweenfactors could also be assessed on the basis of the model. After15 different models were tested, it was concluded that thefactors eliminated did not change the order of importance ofthe remaining factors. When low unstimulated salivary flowrates were included in the model, odds ratios for other factors,especially the consumption of citrus fruits, increased consid-erably. The final aim was to arrive at a logistic multivariablemodel that would facilitate computation of adjusted odds ratios(AOR) and 95% confidence limits for each significant factor(Adena and Wilson, 1982).

For determination of threshold limits for consumption ofcommon citrus fruits and soft drinks, unadjusted odds ratioswere estimated based on five dichotomization values (limitsare given in Table 5).The population-attributable risk (PAR) is defined as the per-

centage decrease in the occurrence of dental erosion in thesource population if the erosion-causing factor is eliminated.Population in this study means the patients of MetropolitanHelsinki dentists. PAR values were calculated from the for-mula a(AOR-1)/[a(AOR-1)+ 1)] where "a" is the occur-rence of exposure in the control group.

Results.The prevalence of dental erosion in patients of dentists in

the greater Helsinki area was 5% (95% confidence intervalfrom 1 to 9%).

Risk factors.-In the 106 cases of erosion, at least one risk

factor was found in 105 patients. In three cases (a typographer,a laboratory technician, and a research assistant), an occupa-tional cause was possible. All were, however, also exposed todietary and gastric risk factors. None of the controls was ex-posed to occupational risk factors. When exposure groups wereformed according to the limits given in Table 4, an acid dietwas associated with erosion in 26 cases, with occurrence ofgastric symptoms in another 26 cases, and with both factorsin 46 cases. Among the remaining eight cases not exceedingthe exposure limits given in Table 4, one had used hydrochloricacid as a medicine, two had very low unstimulated salivaryflow rates and had consumed acid diets for a long time, andfour had eaten citrus fruits or other acidic foodstuffs daily orweekly for many years. In one 61-year-old man with erosion,no risk factor could be identified. When the control patientswere grouped in the same way, dietary exposure was found tobe present in 21, gastric exposure in 19, and both factors infive. Fifty-five of the controls did not exceed any of the ex-posure limits.

Dietary factors. -All of the 106 cases and 98 of the controlshad consumed some acidic food. There was no detectable dif-ference in the duration of consumption of acidic foodstuffs be-tween the two groups, but the intake frequency was much higheramong cases than among controls. Details of the intake frequen-cies and of unadjusted odds ratios are given in Table 1.

Special diets had been eaten by 13 of the erosion cases andseven of the controls. Three cases, but no controls, had eatena weight-reducing diet. Ten cases and seven controls werevegetarian.A subjective feeling of dry mouth was reported by 33 ero-

sion cases and 22 controls. Nine of the cases had sought relieffrom this feeling by consuming acidic drinks. None of thecontrols had done so.

Gastric symptoms.-Among the erosion cases and controls,35 and 12, respectively, had experienced gastric disease di-agnosed by a physician. Seven of the cases were suffering fromanorexia nervosa. None of the controls suffered from this dis-order. The types and frequencies of the various gastric symp-toms, and unadjusted odds ratios, are shown in Table 2.

Saliva.-Salivary flow rates, salivary pH values, and sali-vary buffer capacities are given in Table 3. Male erosion caseshad lower calcium and phosphorus levels in their saliva thandid male controls (Table 3). The stimulated salivary flow ratewas low in 16 cases and six controls. The unstimulated salivaryflow rate was low in seven cases and in six controls.The more important associations with erosion are listed in

Table 4 on the basis of the final multivariable logistic modelarising from the data of the study. In this Table, the adjustedodds ratios are given. No interactions between the various riskfactors were found in the analyses. The population-attributablerisks give percentage decreases in the occurrence of dental

TABLE 1DAILY CONSUMPTION OF ACIDIC PRODUCTS AND

UNADJUSTED ODDS RATIOSDaily Consumers

No. of No. of OddsProduct Cases Controls RatioCitrus fruits 53 (17) 38 (1) 2 (19)Other fruits, berries 42 (11) 29 (0) 2Soft drinks 32 (12) 10 (2) 4 (6)Sport drinks 8 (1) 2 (1) 4 (1)Pickles 4 (0) 0 (0)Apple vinegar 9 (0) 2 (0) 5

Figures in parentheses show those eating the indicated product morethan twice a day and related odds ratios.

943Vol. 70 No. 6

944 JARVINEN et al.

TABLE 2SUBJECTS WITH GASTRIC SYMPTOMS AND UNADJUSTED

ODDS RATIOS

No. of Cases No. of Controls OddsSymptoms Weekly Daily Weekly Daily Ratio*Vomiting 11 5 1 0 18Sour taste 32 8 5 0 12Belching 28 17 7 0 10Heartburn 30 16 15 0 4Stomach-ache 34 1 10 1 4Pain on awakening 8 1 2 0 5

Total# 72 24 7*Odds ratios for those with gastric symptoms weekly or more often.#Numbers of persons with one or more gastric symptoms weekly or

more often.

erosion in the source population, i.e., patients visiting dentistsin the Helsinki district, when the erosive factor was eliminated.Odds ratios for commonly consumed citrus fruits and soft

drinks in relation to the consumption frequencies are shown inTable 5. Because the unadjusted odds ratios increased mark-edly when frequency of consumption of citrus fruits exceededtwice a day, and when the frequency of consumption of softdrinks was once a day or more, these frequencies may be takenas threshold values for erosion.

Discussion.Samples drawn from the population of the Metropolitan Hel-

sinki area were studied. The erosion patients were found byprivate dentists and dental health centers in the area. Dentalerosion may be more common than generally thought, sincethe prevalence in this dental patient population proved to beas high as 5% in the careful examination by one dentist whoapplied strictly defined erosion criteria.The 106 cases and 100 controls were numerous enough to

allow for estimation of the effects of risk factors, despite thefact that the confidence intervals for the odds ratios were rel-atively wide.More subjects were studied than in any previously reported

investigation in which potential causes of erosion were non-industrial. So far, the largest sample size came from a studyin Wales: It consisted of 72 patients referred to the Conserv-ative Department of the Dental Hospital, Cardiff, during aperiod of nine years, and was based on a population of betweenone and two million (Eccles, 1979). When interest in the studyof erosion cases was made public in the Helsinki district, whichhas a population of about 800,000, 106 erosion cases werereferred for examination in one year. This suggests either thatthe prevalence of erosion has increased or that the lesion isbecoming better recognized by dentists.The present study was subject to some potential sources of

error. The case series is likely to present proportionally moresevere forms of dental erosion than in the population at large.First, cases were found in a dental patient population. Second,only 106 erosion cases were found in one year in this largepopulation. When strict criteria were applied to a random sam-ple of 100 patients from the same population, five new caseswere found. Thus, the study may overestimate the relativeimportance of at least some risk factors. On the other hand,the controls were also from the patient population, which re-duces the risk of overestimation.

Another potential source of error is the interview procedure.In many instances, the subject's recollection could have beeninaccurate. However, the same interviewer asked the same

TABLE 3STIMULATED AND UNSTIMULATED SALIVARY FLOW RATES,pH, BUFFERING CAPACITY, CALCIUM, PHOSPHORUS, AND

PROTEIN LEVELS

Erosion Cases ControlsMean SD Mean SD

MenSalivary secretion (mlmin)

Unstimulated 0.6 0.3 0.5 0.3Stimulated 2.1 1.1 2.1 1.1

pHUnstimulated 7.2 0.3 7.1 0.5Stimulated 7.4 0.2 7.3 0.2

Buffering capacity 5.3 1.1 5.4 1.0Ca unstim. (mmol/L) 0.9 0.3** 1.3 1.0Ca stim. " 0.8 0.3*** 1.2 0.5P unstim. " 4.1 1.5** 5.2 2.1P stim. 2.4 1.0** 3.0 1.0Proteins, unstim. (mg/mL) 1.5 0.5* 1.8 0.7Proteins, stim. " 1.2 0.4 1.3 0.4WomenSalivary secretion

Unstimulated 0.5 0.9 0.5 0.3Stimulated 1.7 0.9 1.7 0.7

pHUnstimulated 7.1 0.3 7.1 0.3Stimulated 7.4 0.2 7.4 0.2

Buffering capacity 5.2 1.2 5.0 1.0Ca unstim. (mmol/L) 1.1 0.5 1.2 0.7Ca stim. " 0.9 0.3* 1.0 0.4P unstim. " 5.4 2.5 4.7 1.9P stim. " 2.6 1.2 2.9 1.2Proteins, unstim. (mg/mL) 1.7 0.8 1.5 0.8Proteins, stim. N 1.1 0.5 1.2 0.5

Significance of differences between means: *p<0.05; **p<0.01;***p<0.001.

questions of all 106 cases and 100 controls, using a standardform. Note that none of the cases had been interviewed beforeregarding the possible causes of erosion. If answers were in-accurate, they were likely to have been similarly so for bothcases and controls.The study findings suggest that the most likely cause of

dental erosion can normally be found if a patient's history istaken carefully. The existence of idiopathic erosion, with noobvious cause, is doubtful. In only one instance could a causeof erosion not be identified.

Extrinsic and intrinsic causes of erosion were equally com-mon. Each was observed alone in 26 instances. In a previousstudy, simultaneous exposure to both extrinsic and intrinsiccauses of erosion was recorded in 10% of cases (Eccles andJenkins, 1974). In the study reported here, the two types ofrisk factor were detected simultaneously in 46 cases, Le., abouttwice as often as either alone. It is important for each patient'shistory to be taken with respect to both dietary and gastricfactors, so that the overlooking of any significant exposure canbe avoided. On the other hand, occupational exposure is nowrare. In this study, three instances of possible occupationalexposure were seen. In only one of them, a typographer, wasthe severe erosion detected likely to have been caused by thesulfuric and nitric acid fumes to which the patient had beenexposed for 15 years in the 1940's and 1950's.

Demineralizing acidic foods commonly consumed by theerosion cases were citrus fruits, soft drinks, sport drinks, andapple vinegar. When citrus fruits were consumed more thantwice a day, this was associated with an erosion risk 37 timesgreater than that in those who consumed citrus fruits less often.It also appears risky to consume apple vinegar or sport drinks

J Dent Res June 1991

RISK FACTORS IN DENTAL EROSION

TABLE 4FACTORS ASSOCIATED WITH DENTAL EROSION, ADJUSTEDODDS RATIOS, AND POPULATION-ATTRIBUTABLE RISK

Adjusted 95% Population-Odds Confidence attributable

Factor Ratio Interval Risk (%)Citrus fruits 37 4-369 26

(more than twice a day)Vomiting 31 3-300 23

(weekly or more often)Other gastric symptoms 10 4-22 67

(weekly or more often)Apple vinegar 10 2-57 15

(weekly or more often)Soft drinks 4 2-10 26

(four to six or more per week)Sport drinks 4 1-14 15

(weekly or more often)Saliva unstim. 5 1-18 19

(50.1 mL/min)

once a week or more often, or soft drinks daily. These habitsresulted in the risk of erosion being ten, four, and four timesgreater, respectively, than when the habit did not exist.

Although it has been known for a long time that acids cancause dental erosion, the frequency of exposure and the mag-nitude of the risks have remained essentially unknown. Theodds ratios reported here indicate threshold values for com-monly consumed acidic dietary components, such as citrusfruits and soft drinks, which resulted in a marked increase inerosion risk. For citrus fruits, for instance, the critical fre-quency of consumption was more than twice a day. For softdrinks, it was once a day or more.Many previous reports have indicated that dental erosion can

be caused by the eating of citrus fruits (Miller, 1907; Pickerill,1926; Eccles and Jenkins, 1974; Linkosalo and Markkanen,1985; Smith and Shaw, 1987). Demineralization by citric acidhas also been shown in vitro (Miller, 1952; Davis and Winter,1977; Linkosalo et al., 1988; Ryt6maa et al., 1988). Citrusfruits and their juices are strongly acidic (pH less than 4.0)(Miller, 1952; Touyz and Glassman, 1981; Meurman et al.,1987). The demineralizing effect of citric acid is exceptionallygreat because its chelating action on enamel calcium continueseven after the pH increases at the tooth surface (McClure andRuzicka, 1946; Elsbury, 1952). Very frequent consumption ofcitrus fruit juices can affect the enamel, even over a short term.For instance, 350 g of grapefruit juice daily for four weeksproduced detectable changes in the enamel surface (Thomas,1957).Many soft drinks contain citric, phosphoric, carbonic, and

other acids. Their pH value is often less than 4.0 (Eccles andJenkins, 1974; Touyz and Glassman, 1981). Soft drinks, ex-cept those containing just carbonic acid, have been reported tocause dental erosion, both in case studies (Eccles and Jenkins,1974; High, 1977; Mueninghoff and Johnson, 1982) and invitro (Ryt6maa et al., 1988). Over the last 20 years, sales ofsoft drinks in Finland have increased three-fold (Pallari, 1988).Among those taking strenuous physical exercise, special drinksare commonly consumed for the maintenance of fluid balanceand the provision of carbohydrates for energy. In studies inanimals and in vitro, such sport drinks have turned out to bestrongly erosive (Rytomaa et al., 1988; Sorvari and Kiviranta,1988; Sorvari, 1989). The most important erosive componentin these drinks is citric acid, included for its refreshing taste.However, citric acid could be replaced by the less erosivemalic acid (Meurman et al., 1990). As far as we know, noerosive effect of apple vinegar has previously been reported.

TABLE 5UNADJUSTED ODDS RATIOS IN RELATION TO FREQUENCY OF

CONSUMPTION OF CITRUS FRUITS AND SOFT DRINKS

95%Confidence

Consumption Odds Ratio IntervalCitrus fruitsnone 0.8 0-11-3 times a week 0.8 0-24-7 times a week 1.2 1-32 times a day 1.1 0-3>2 times a day 19.0 2-155Soft drinksnone 0.6 0-11-3 times a week 1.7 0-34-7 times a week 3.5 1-82 times a day 14.2 2-109>2 times a day 9.5 2-44

However, it has been observed that frequent consumption ofpickles causes erosion in those eating lactovegetarian diets(Linkosalo and Markkanen, 1985).

In the study reported here, vomiting and other gastric dis-orders were important risk factors. In the logistic model, theadjusted odds ratios varied between gastric symptoms. Vom-iting once a week or more was associated with an erosion risk31 times greater than that in patients who vomited less fre-quently than once a week. The other relatively common gastricsymptoms-i.e., acid taste in the mouth, belching, heartburn,stomach-ache, and gastric pain on awakening-were associ-ated with an erosion risk ten times greater when the frequencywas once a week or more often, than when it was less. In allof these disorders, the origin of the acid is the same, but itsconsistency and adherence to the surface of the tongue differ.The intensity of a symptom is not necessarily associated withthe severity of a disease. A morphologically mild conditioncan cause many symptoms (Salo, 1986). A patient may ignorethe gastric symptoms of a minor disease, considering themnormal.

Salivary calcium and phosphorus levels were lower in theerosion cases than in the controls, but the logistic model in-dicated that these factors were not associated with erosion. Thelow levels of calcium and phosphorus in the saliva must, there-fore, be related to some other risk factor, such as a low un-stimulated salivary flow rate. The study showed that this latterfactor was important. A patient with an unstimulated salivaryflow rate of 0.1 mL/min or less was at five-times-greater riskof erosion than those with higher flow rates. Other authorshave also pointed out that the unstimulated salivary flow rateis an important factor determining whether dental erosion oc-curs (Hellstrim, 1977; Woltgens et al., 1985). At normal sal-ivary flow rates, acidic drinks are eliminated from the mouthin about ten min, and the pH at the tip of the tongue remainslow for only some two min after the drink has been consumed(Meurman et al., 1987). In contrast, in patients with low sal-ivary flow rates, the pH remains low for about 30 min (Ten-ovuo and Rekola, 1977). In this study, a low unstimulatedsalivary flow rate was associated with dental erosion even whenthe frequency of consumption of acidic foods was below thethreshold limit. One patient, who did not secrete any unstim-ulated saliva, had eaten citrus fruit twice a week to stimulatesalivary flow. Another patient had eaten citrus fruits weeklyand other fruits daily. Salivary flow rate is rarely measured bydentists (Telivuo and Murtomaa, 1988), but the unstimulatedsalivary flow rate is particularly important because it is the rateat which saliva flows for most of the day (Sreebny, 1989).The population-attributable risk of this study gives the over-

Vol. 70 No. 6 945

946 JARVINEN et al.

all risk of dental erosion in the dental patient population. Al-though the relative effects of citrus fruits and vomiting on therisk of erosion were strong (AOR 37 and 31, respectively),the absolute decrease in the incidence of erosion would befairly modest, because the proportion of dental patients fre-quently exposed to these factors was small. In contrast, manygastric factors are common among the population, and theirrelatively weak erosive effects (AOR 10) were associated witha high population-attributable risk (67%). In other words, elim-ination of these factors would reduce the prevalence of dentalerosion. Apple vinegar was twice as powerful an erosion-caus-ing factor as soft drinks, but, because it is not commonlydrunk, its PAR is small (15%). The popularity of soft drinksis such that they are associated with a PAR of 26%, compa-rable with the level for citrus fruits, although the latter aremuch more erosive.The present study illustrates the importance of non-industrial

risk factors in dental erosion. Many such factors could be elim-inated by general measures, such as increasing the availabilityof information about acidic products and gastric conditions,and through product development. It is also important for ero-sion to be diagnosed at an early stage and the risk factors tobe identified. This increases the possibilities of successfultreatment and reduces complications associated with mechan-ical intervention.

Acknowledgment.We are most grateful to Mr. Kari Hdrnninen, M.Sc., for his

contribution to the data analysis.

REFERENCES

ADENA, M.A. and WILSON, S.R. (1982): Case-control Studies. In:Generalised Linear Models in Epidemiological Research, Syd-ney: The Instat Foundation, Chap. 6, pp. 1-13.

ASHER, C. and READ, M.J.F. (1987): Early Enamel Erosion inChildren Associated with the Excessive Consumption of CitricAcid, Br Dent J 162:384-387.

CENTERWALL, B.S.; ARMSTRONG, C.W.; FUNKHOUSER, L.;and ELZAY, R. (1986): Erosion of Dental Enamel among Com-petitive Swimmers at a Gas-chlorinated Swimming Pool, Am JEpidemiol 123:641-647.

CLARK, D.C. (1985): Oral Complications of Anorexia Nervosa and/or Bulimia: With a Review of the Literature, J Oral Med 40:134-138.

DAWES, C. (1970): Effects of Diet on Salivary Secretion and Com-position, J Dent Res 49:1263-1273.

DAVIS, W.B. and WINTER, P.J. (1977): Dietary Erosion of AdultDentine and Enamel, Br Dent J 143:116-119.

ECCLES, J.D. (1978): Erosion of Teeth by Gastric Contents, Lancet2:479.

ECCLES, J.D. (1979): Dental Erosion of Nonindustrial Origin. AClinical Survey and Classification, J Prosthet Dent 42:649-653.

ECCLES, J.D. and JENKINS, W.G. (1974): Dental Erosion and Diet,J Dent 2:153-159.

ELSBURY, W.B. (1952): Hydrogen-ion Concentration and Acid Ero-sion of the Teeth, Br Dent J 93:177-179.

ERIKSSON, J.-H. and ANGMAR-MANSSON, B. (1986): Ero-sionsskador av C-Vitamintuggtabletter, Tandldkartidningen 78:541-544.

FISKE, C.H. and SUBBAROW, Y. (1925): The Colorimetric Deter-mination of Phosphorus, J Biol Chem 66:375-400.

FROSTELL, G. (1973): Betydelsen ur Cariologisk Synpunkt av Stbr-ningar i Salivsekretionen, Swed Dent J 66:69-80.

HELLSTROM, I. (1977): Oral Complications in Anorexia Nervosa,Scand J Dent Res 85:71-86.

HIGH, A.S. (1977): An Unusual Pattern of Dental Erosion. A CaseReport, Br Dent J 143:403-404.

HURST, P.S.; LACEY, L.H.; and CRISP, A.H. (1977): Teeth,

Vomiting and Diet: A Study of the Dental Characteristics of Sev-enteen Anorexia Nervosa Patients, Postgrad Med J 53:298-305.

ISMAIL-BEIGI, F.; HORTON, P.F.; and POPE, C.E., II (1970):Histological Consequences of Gastroesophageal Reflux in Men,Gastroenterology 58:163-174.

JAMES, P.M.C. and PARFITT, G.J. (1953): Local Effects of CertainMedicaments on the Teeth, Br Med J 2:1252-1253.

JARVINEN, V.; MEURMAN, J.H.; HYVARINEN, H.; and RY-TOMAA, I. (1988): Dental Erosion and Upper GastrointestinalDisorders, Oral Surg Oral Med Oral Pathol 65:298-303.

KNEWITZ, J.L. and DRISKO, C.L. (1988): Anorexia Nervosa andBulimia: A Review, Compend Contin Educ Dent 9:244-247.

LINKOSALO, E. and MARKKANEN, H. (1985): Dental Erosionsin Relation to Lactovegetarian Diet, Scand J Dent Res 93:436-441.

LINKOSALO, E.; MARKKANEN, S.; ALAKUIJALA, A.; andSEPPA, L. (1988): Effects of Some Commercial Health Bever-ages, Effervescent Vitamin C Preparations and Berries on HumanDental Enamel, Proc Finn Dent Soc 84:31-38.

LOWRY, O.H.; ROSEBROUGH, N.J.; FARR, A.L.; and RAN-DALL, R.J. (1951): Protein Measurement with the Folin PhenolReagent, J Biol Chem 193:265-275.

LYNCH, J.B. and BELL, J. (1947): Dental Erosion in Workers Ex-posed to Inorganic Acid Fumes, Br J Ind Med 4:84-86.

MALCOLM, D. and PAUL, E. (1961): Erosion of the Teeth due toSulphuric Acid in the Battery Industry, Br J Ind Med 18:63-69.

MANNERBERG, F. (1963): Saliva Factors in Cases of Erosion, OdontRevy 14:156-166.

McCLURE, F.J. and RUZICKA, S.J. (1946): The Destructive Effectof Citrate vs. Lactate Ions on Rats' Molar Tooth Surfaces, in vivo,JDentRes 25:1-12.

MEURMAN, J.H. and MURTOMAA, H. (1986): Effect of Effer-vescent Vitamin C Preparations on Bovine Teeth and on SomeClinical and Salivary Parameters in Man, Scand J Dent Res 94:491-499.

MEURMAN, J.; RYTOMAA, I.; KARI, K.; LAAKSO, T.; andMURTOMAA, H. (1987): Salivary pH and Glucose after Con-suming Various Beverages, Including Sugar-containing Drinks,Caries Res 21:353-359.

MEURMAN, J.H.; HARKONEN, M.; NAVERI, H.; KOSKINEN,J.; TORKKO, H.; RYTOMAA, I.; JARVINEN, V.; and TU-RUNEN, R. (1990): Experimental Sports Drinks with MinimalDental Erosion Effect, Scand J Dent Res 98:120-128.

MILLER, C.D. (1952): Enamel Erosive Properties of Fruits and Var-ious Beverages, JAm Diet Assoc 28:319-324.

MILLER, W.D. (1907): Experiments and Observations on the Wast-ing of Tooth Tissue Variously Designated as Erosion, Abrasion,Chemical Abrasion, Denudation, etc., Dent Cosmos 49:225-247.

MUENINGHOFF, L.A. and JOHNSON, M.H. (1982): Erosion: aCase Caused by Unusual Diet, JAm Dent Assoc 104:51-52.

MYLLARNIEMI, H. and SAARIO, I. (1985): A New Type of Sli-ding Hiatus Hernia, Ann Surg 202:159-161.

NEGUS, T.W. and TODD, J.O. (1986): Bulimia Nervosa in a Male,Br Dent J 160:290-291.

NELDER, J.A. and WEDDERBURN, R.W.M. (1972): GeneralizedLinear Models, J R Statist Soc A 135:370-384.

PALLARI, M. (1988): Mita Suomessa Juodaan. In: Juomat '88 Tut-kimusten Valossa, Helsinki: AC-Tiedotus, pp. 6-11.

PICKERILL, H.P. (1926): Verhutung von Zahnkaries und Mund-sepsis, Berlin: Herman Meusser, pp. 131-139.

PINDBORG, J.J. (1970): Pathology of Dental Hard Tissues, Co-penhagen: Munksgaard, pp. 312-321.

POPE, C.E., II (1982): Gastroesophageal Reflux Disease. In: CecilTextbook of Medicine, J.B. Wyngaarden and L.H. Smith, Jr.,Eds., Philadelphia: Saunders Co., pp. 623-624.

RYTOMAA, I.; MEURMAN, J.H.; KOSKINEN, J.; LAAKSO, T.;GHARAZI, L.; and TURUNEN, R. (1988): In vitro Erosion ofBovine Enamel Caused by Acidic Drinks and Other Foodstuffs,Scand J Dent Res 96:324-333.

SALO, J. (1986): Reflux Esophagitis and its Treatment, Duodecim102:1395-1398.

SKOGEDAL, O.; SILNESS, J.; TANGERUD, T.; LAEGREID, O.;and GILHUUS-MOE, 0. (1977): Pilot Study on Dental Erosion

J Dent Res June 1991

Vol. 70 No. 6 RISK FACTORS IN DI

in a Norwegian Electrolytic Zinc Factory, Community Dent OralEpidemiol 5:248-251.

SMITH, A.J. and SHAW, L. (1987): Baby Fruit Juices and ToothErosion, Br Dent J 162:65-67.

SMITH, B.G.N. and KNIGHT, J.K. (1984): A Comparison of Pat-terns of Tooth Wear with Aetiological Factors, Br DentJ 157:16-19.

SORVARI, R. (1989): Effects of Various Sport Drink Modificationson Dental Caries and Erosion in Rats with Controlled Eating andDrinking Pattern, Proc Finn Dent Soc 85:13-20.

SORVARI, R. and KIVIRANTA, I. (1988): A Semiquantitative Methodof Recording Experimental Tooth Erosion and Estimating OcclusalWear in the Rat, Arch Oral Biol 33:217-220.

SREEBNY, L.M. (1989): Recognition and Treatment of Salivary In-duced Conditions, Int Dent J 39:197-204.

TELIVUO, M. and MURTOMAA, H. (1988): Preventive Dentistryin Private Practice in Finland, Proc Finn Dent Soc 84:183-190.

ENTAL EROSION 947

TEN BRUGGEN CATE, H.J. (1968): Dental Erosion in Industry, BrJ Ind Med 25:249-266.

TENOVUO, J. and REKOLA, M. (1977): Some Effects of Sugar-flavoured Acid Beverages on the Biochemistry of Human WholeSaliva and Dental Plaque, Acta Odontol Scand 35:317-330.

THOMAS, A.E. (1957): Further Observations on the Influence ofCitrus Fruit Juices on Human Teeth, NY State DentJ 23:424-430.

TOUYZ, L.Z.C. and GLASSMAN, K.M. (1981): Citrus, Acid andTeeth, Tydskr Tandheelkd Ver S Afr 36:195-201.

TRUDEAU, D.L. and FREIER, E.F. (1967): Determination of Cal-cium in Urine and Serum by Atomic Absorption Spectrophotom-etry (AAS), Clin Chem 13:101-114.

WOLTGENS, J.M.H.; VINGERLING, P.; DE BLIECK-HOGER-VORST, J.M.A.; and BERVOETS, D.J. (1985): Enamel Erosionand Saliva, Clin Prev Dent 7:8-10.